Scientific MethodIdentifying HypothesesIntroduction to ResearchLecture 11

Key Lecture ConceptsUnderstanding the process described as the scientific methodThe role of a hypothesis in a research studyStrategies underlying hypothesis formulationThe manner to frame your statement of a hypothesis

*

Research is .Knowledge acquisition gained through reasoningthrough intuitionbut most importantly through the use of appropriate methodsThe Scientific Method*

Basic Elements of the Scientific MethodEmpiricism: the notion that enquiry is conducted through observation and knowledge verified through evidenceDeterminism: the notion that events occur according to regular laws and causes. The goal of research is to discover theseScepticism: the notion that any proposition is open to analysis and critique*

Scientific MethodChoose a question to investigateIdentify a hypothesis related to the questionMake testable predictions in the hypothesisDesign an experiment to answer hypothesis questionCollect data in experimentDetermine results and assess their validityDetermine if results support or refute your hypothesis*

The Scientific Method1. Suspicion that a factor (exposure) may influence occurrence of disease or a noted health outcome

Observations in clinical practiceExamination of disease/outcome patternsDo subpopulations have higher or lower rates?Are disease rates increased in the presence of certain factors?Observations in laboratory researchTheoretical speculation*

The Scientific Method2. Identify variables you are interested in:Exposure - (risk factor, protective factor,predictor variable, treatment) Outcome - (disease, event)

3. Formulate a specific hypothesisFrame a hypothesis which seeks to answer a specific question about the relationship between an exposure and an outcome*

Basic Question in Research Are exposure and disease/outcome linked?ExposureDisease / Health OutcomeIs there an association between them?*

Next Step: Design Study Study Designs (not exhaustive)Case seriesCross-sectionalCase-controlCohortRandomized controlled clinical trial*

Association From the results of your study, does a statistical relationship exist between two or more events, characteristics, or other variables

Is there a statistical relationship, or association, between exposure and disease/outcome?

*

Statistical Association

The degree to which the rate of diseaseor outcome in persons with a specific exposure is either higher or lower than the rate of disease or outcome among those without that exposure.*

The Scientific MethodAssess validity of association

Does the observed association really exist?Is the association valid?Are there alternative explanations for the association?ChanceBiasConfounding*

HypothesesShape and guide a research study in terms of:

identification of study sample sizewhat issues should be involved in data collectionthe proper analysis of the datadata interpretation

*

Hypothesis Formulation

---Formulate a hypothesis

---Frame the hypothesis ina format that is testable

---Test the hypothesis*

Hypothesis FormulationObservations from:

Literature (review PubMed on topic area)Natural experiments (e.g. migrant studies)Multi-national comparisonsDescriptive studies (assessment of person, place, and time characteristics)Creativity*

*Cervical Cancer

Infectious and chronic diseases show great variation from one country to another.

Some differences may be attributed to:

---Climate---Cultural factors---Diet---Genetics*

Descriptive Study DesignsUsed to help formulate hypotheses*

Case Series ApproachIdentify the experience of a group of patients with a similar diagnosis, orIdentify the experience of a group of individuals with an exposure in common

Patients or individuals may be identified from a single or multiple sources

*

Population Survey ApproachDescribe issues related to disease or exposure in populationsUsually rely upon routinely collected data from established surveillance or notifiable disease systems*Unique Component: usually identify the characteristics of an issue from a representative sample of the population

Three essential characteristics that we look to measure in descriptive studies are...PersonPlace Time*

PersonSince disease not does occur at random:

What kinds of people tend to develop a particular disease, and who tends to be spared? Whats unusual about those people?*

Person FactorsAge, gender, race, ethnicityGenetic predispositionConcurrent diseaseDiet, exercise, smokingRisk taking behaviorSES, education, occupation*

Place

Since disease not does occur at random:

Where is the disease especially common or rare, and what is different about those places?*

Place FactorsGeographic placeresidenceoccupationclimategeologypopulation densityeconomic developmentnutritional practicesmedical practices*

TimeSince disease not does occur at random:

How does disease frequency change over time, and what other factors are temporally associated with those changes?*

Time FactorsCalendar Time / Time of DayTime since an eventDate of onsetAge (time since birth in the young)SeasonalityTemporal trends*

Remember the Elements of the Scientific MethodDiscoveries or hypotheses are sometimes resisted because they seem counter-intuitive*

Hypothesis FramingTraditionally..

H0:Null hypothesis (assumed)H1:Alternative hypothesis

*

Case Series (in practice)Description of clinical/epidemiologic characteristics of a number of patients with a given disease- usually a consecutive set of clinical cases of disease (or health issue)

Analyze cases together to learn about the disease (be careful as results do not demonstrate temporal relationships)*Mandil

H0:There is no association between theexposure and disease of interest

H1:There is an association between theexposure and disease of interest(beyond what might be expected from random error alone)*Hypothesis Framing

Another Type of Framing:

What is the best estimate of the risk of disease in those who are exposed compared to those who are unexposed (i.e. exposed are at XX times higher risk of disease).

This moves away from the simple dichotomy of yes or no for an exposure/disease association to the estimated magnitude of effect irrespective of whether it differs from the null hypothesis.*Hypothesis Framing

Ways to Express Hypotheses:

1.Suggest possible events

The rate of survival will increase after surgery.*Hypothesis Framing

Ways to Express Hypotheses:

2.Suggest relationship between specificexposure and health-related event

A high cholesterol intake is associatedwith the development (risk) of coronaryheart disease. *Hypothesis Framing

Ways to Express Hypotheses:

3.Suggest cause-effect relationship.

Cigarette smoking is a cause of lung cancer*Hypothesis Framing

Ways to Express Hypotheses:

4.One-sided vs. Two-sided

One-sided example:Helicobacter pylori infection is associatedwith increased risk of stomach ulcer

Two-sided example:Weight-lifting is associated with risk oflower back injury*Hypothesis Framing

Guidelines for Framing Hypotheses:

State the exposure to be measured as specifically as possible.

State the health outcome as specifically as possible.

Strive to explain the smallest amount of ignorance*Hypothesis Framing

Example Hypotheses:

POOREating junk food is associated with the development of cancer.

GOODThe human papilloma virus (HPV) subtype 16 is associated with the development of cervical cancer.*Hypothesis Framing

The Next StepFormally test the identified hypotheses in a research studyThe study should follow a specific plan or protocol (the study design)Study designs direct how the investigation is conducted and allows for the translation of a conceptual hypothesis into an operational one*Mandil

NewtonDisappointment is when a beautiful hypothesis is destroyed by an ugly fact*

**The fundamental concepts of this lecture are outlined here. We will begin by discussing the distinction between two types of epidemiology; descriptive epidemiology and analytical epidemiology. The second one-half of the class will focus to a large extent on the on concept of analytical epidemiology. This lecture begins by highlights what analytical epidemiology is and how it is distinguished from descriptive epidemiology. A fundamental concept in analytical epidemiology is the exposure and disease model. This model will be re-emphasized here.

Next, the lecture turns to specific measures of association that are used in epidemiology. The two major measures are the relative risk and odds ratio measures. The fundamental characteristics of these measures will be distinguished, and an illustration of their computation will be provided.